basic 2 anisotropic wet-etching of silicon: characterization and

Basic 2Basic 2Anisotropic Wet-etching

of Silicon: Characterization and Modeling of ChangeableModeling of Changeable

Anisotropypy

Prof K SatoProf. K. SatoDept. of Micro/Nano Systems Engineering

Nagoya University

Basic 2 Anisotropic Wet-etching of Silicon: Prof. K. SatoCharacterization and Modeling of Changeable AnisotropyCOE for Education and Research of Micro-Nano Mechatronics, Nagoya University

Orientation Dependent Etching(Conventional Products)

〈110〉

〈111〉

〈112〉

〈100〉 SiO2〈111〉

〈100〉 SiO2

Si

Diaphragm

Deep grooves on a (110) wafer Diaphragm on a (100) wafer


Variation in etching profile on (100) silicon wafer

Groove Wall Orientations(111) (100) (110)(111) (100) (110)

KOHTMAH

KOHTMAH

KOH/IPATMAH/ f t tTMAH

EDPTMAH TMAH/surfactant

EDP


Variation in etching profile on (110) silicon wafer

Groove Wall OrientationsGroove Wall Orientations(111) (111) (100) (110)

KOHTMAHEDP

KOHTMAHEDP

KOHTMAH

KOH/IPATMAH/surfactantEDPEDP EDP EDP


Non-conventional 3-D microstructures using KOH anisotropic etching

Etching from both sides of a wafer Two step etching using two mask layers Etching from both sides of a wafer Two-step etching using two mask layers

K. Sato, et al, Proc. IFToMM Intl. Micromechanism Symp. (Tokyo, 1993.6) 155-160


Densely Arrayed Silicon Needles with a Pitch Distance of 200 microns Aiming at Transdermal Drug Delivery

M Shikida et al :Proc MEMS 03 (Kyoto 2003) 562M. Shikida et al.:Proc. MEMS-03 (Kyoto, 2003), 562


New types of anisotropically etched 3-D structures: Curved, Sharp-cornered, 45-degree-angled V-grooves

Prem Pal: Jpn. J. Appl. Phys. 49 (2010)056702

[ ][100]

[110]

[110]


Anisotropic chemical etching of Sifrom MEMS Point of View

• Etching SolutionsKOH, “TMAH”, “EDP”, N2H4, NaOH, CsOH, etc.

Ch i l ti• Chemical reactionSi + 2OH- + 2H2O

Si(OH) + H SiO (OH) 2 + 2H→ Si(OH)4 + H2 → SiO2(OH)22- + 2H2

• What are known;Si (111) shows an extremely low etch rateSi (111) shows an extremely low etch rate.Etch-stop techniques: B-dope, Electro-chemical,

etc.etc.• Applications

Diaphragms, V-grooves, CantileversDiaphragms, V grooves, Cantilevers ---Limitations in fabricated shapes

Many mysteriesBasic 2 Anisotropic Wet-etching of Silicon: Prof. K. SatoCharacterization and Modeling of Changeable AnisotropyCOE for Education and Research of Micro-Nano Mechatronics, Nagoya University

Many mysteries

Sequential reactions of Si etching in alkaline solution

Si + 2OH- + 2H2O → SiO2(OH)22- + 2H2

f fThis is the results of the following steps.(R. A. Wind, M.A. Hines, Surface Science 460 (2000) 21-38)

＝SiH2 + OH- + H2O → ＝SiHOH+ H2 + OH-2 2 2

＝SiHOH + OH- + H2O → ＝Si(OH)2+ H2 + OH-

( Si) Si(OH) ２H O ２( Si H) Si(OH)(≡Si)2＝Si(OH)2 + ２H2O → ２(＝Si-H)＋Si(OH)4

Si(OH)4 + ２OH- → ＝SiO2 (OH)22-+ ２H2 O( )4 2 ( )2 2


Contents

Characterization of Anisotropic EtchingCharacterization of Anisotropic Etchingin macroscopic domains

“Dangling-Bond Model” does not tell the truth,because no dynamics included. y

“Step Flow Model” explains anisotropy in the vicinity of Si (111)vicinity of Si (111)

*R d i t b t KOH d TMAH*Reversed anisotropy between KOH and TMAH*Etched shape clearly reflects atomic-step p y p

behavior in the vicinity of Si (111)


Contents


“D li B d M d l” d t t ll th t th“Dangling-Bond Model” does not tell the truth,because no dynamics included.

“Step Flow Model” explains anisotropy in the vicinity of Si (111) c ty o S ( )

*Reversed anisotropy between KOH and TMAH*Reversed anisotropy between KOH and TMAH*Etched shape clearly reflects atomic-step



Etching rate measurementK. Sato et al.: Sensors and Actuators A-64 (1998) 87-93.

(110) contact probe(110) contact probe

R = 22 mmprobing network

R 22 mm

R

(001)( )

(110)-

Hemispherical specimen of single-crystal silicon


Hemispherical specimen

Before After

• Maximum etching depth: 100 - 150 µm


Etching rate contour map for a KOH solution

(100)K. Sato et al.: Sensors and Actuators A-64 (1998) 87-93.

90゜ (111)(100)

60゜(110)

30゜

0゜

0゜

30

(110) 90゜

0゜0゜ 30゜ 60゜ 90゜

(100)NAGOYA UNIVERSITY

(100) min max


Effects of a surfactant added to TMAH solutionK. Sato et al.: Sensors and Materials 13-5 (2001) 285-291.

90°

(010)

(110)90°

(010)

(110)(111)

60°

(110)

60°

(110)

30° 30°

0°0° 30° 60° 90°

(100)(001)

0°0° 30° 60° 90°

(100)(001)0 30 60 90(001) 0 30 60 90(001)

min max

25%TMAH + NCW 25%TMAH


Poly-oxiethylene-alkyl-phenyl-ether

C9H19 O(CH2CH2O) HC9H19 O(CH2CH2O)nH

Hydrophobic HydrophilicHydrophobic Hydrophilic

• Liquid easy to operate with little foamingS bl b h i id d lk li l i• Stable both in acid and alkaline solutions


Orientation-dependent effects of surfactant decreasing etch rates of silicon

2.500}2.000

2.500

25%TMAH (without surfactant)k}

/{10

0}

1.500

of {

i j k

1.000 25%TMAH+2%NCW

te ra

tio o

0.500

Etch

rat

0.000{100} {310} {210} {530} {320} {540} {110} {331} {221} {111} {211} {311} {100}

OrientationsOrientations

K. Sato, et al., Sensors and Materials 13-5 (2001) 285-291.


Mask-corner undercut suppression

Effects of the surfactant NCWK. Sato, et al., Sensors and Materials 13-5 (2001) 285-291.

Orientations appearing on a vertex:(111) - ( j j 1 ) - (110)

25 wt.% TMAH, 80 ºCEtching depth: 50 µm

without NCW{111}

{100}

i h{110}

{110}

with 2 wt.% NCW

300Basic 2 Anisotropic Wet-etching of Silicon: Prof. K. SatoCharacterization and Modeling of Changeable AnisotropyCOE for Education and Research of Micro-Nano Mechatronics, Nagoya University

300 µm

Comparison of Structure Shape Etched from Same Mask Apertures

<110>

<110>

Prem Pal, J. Micromech. Microeng. 17 (2007) 2299–2307

Etch depth = 35 μm<110>

( ) (b) ( ) (d)(a) (b) (c) (d)

Pure 25 wt% TMAH

(a) (b) (c) (d)

T-Shaped cantilever

Circular island

Cross island

Cross aperture

25 wt% TMAH + NC-200


cantilever island island aperture

Etching ratedatabase


Anisotropic etching simulation system MICROCAD


Simulation results using MICROCAD

K Sato et al Electronics and Communications in Japan Part2 83 4 (2000)


K. Sato, et al., Electronics and Communications in Japan, Part2, 83-4 (2000).

Simulation results using MICROCAD


Densely Arrayed Silicon Needles with a Pitch Distance of 200 microns for Transdermal Drug Delivery

M. Shikida et al. Sensors and Actuators A 116 (2004) 264–271


Arrayed Needle Fabrication Process:Combination of mechanical dicing and wet etching

M. Shikida et al. J. Micromech. Microeng. 14 (2004) 1462–1467

Etching time: (a) 0 min. (b) 10 min. (c) 20 min.

(d) 30 i ( ) 40 iBasic 2 Anisotropic Wet-etching of Silicon: Prof. K. SatoCharacterization and Modeling of Changeable AnisotropyCOE for Education and Research of Micro-Nano Mechatronics, Nagoya University

(d) 30 min. (e) 40 min.

MICROCAD Simulation Results.Etching time: (a) 0 min., (b) 5 min., (c) 10 min.,

(d) 15 min., (e) 20 min., (f) 25 min.Et hi diti 34 0 t % KOH 80oC

(a) (b) (c)

Etching conditions: 34.0 wt.% KOH, 80oC

(a) (b) (c)

(d) (e) (f)

M Shikida et al J Micromech Microeng 14 (2004) 1462 1467


M. Shikida, et al, J. Micromech. Microeng. 14 (2004) 1462–1467

Etched deep grooves on (110) Si using a KOH water solution

Groove depth: 90 microns


KOH and TMAH show different types of anisotropy

K. Sato, et al.: Sensors and Actuators A-73 (1999) 131-137.

TMAH: 25%, 90°C KOH: 40%, 90°C

90゜90°(010)

60゜60°(430)(110)

30゜30°

(011)

0゜0° (100) 00゜ 30゜ 60゜ 90゜

00° 30° 60° 90°(001) (101)

( 00)


Relation between the etching rate distribution and etched profile

Etching rate distribution

1 0

1.51.0

m/m

in)

m/m

in)

0.5

1.0

0.5

Etc

h ra

te (µ

m

Etc

h ra

te (µ

m

-90° 90°0°0

(111) (111)(110)-90° 90°0°0

(111) (111)(110)(311) (311)

(111)

Etch rate vectorEtched profile analyzed using Wulff-Jaccodine method

(110)

Si

25% TMAH (86°C) 40% KOH (70°C)


Contents



“Step Flow Model” explains anisotropy in the vicinity of Si (111)c ty o S ( )




Conventional Explanation of Anisotropy in Etching

Hypothesis: Number of dangling bond appearing on the silicon surface determines the etching ratesilicon surface determines the etching rate

zz

zxy x y

(100)面(110)面

Dangling bond：2Dangling bond：1 +(2)（ d b k b d ）

xy

Dangling bond：2Back bond：2

（Exposed back bond：2）Back bond：3

(111)面Dangling bond：1Back bond：3


Step Flow Model for (111) SiliconM. Elwenspoek, J. Electrochem. Soc. 140-7 (1993) 2075-80


Two types of stable steps on Si (111) surface(Mono-hydride and Di-hydride steps)

M A Gosalvez et al J Micromech Microeng 17 (2007) S1 S26


M.A. Gosalvez, et al. J. Micromech. Microeng. 17 (2007) S1–S26

Step movements determine profiles ofpits and mesa on (111) silicon surface


Atomic level etching simulation in the vicinity of Si (111)

(111)(110)

M.A. Gosalvez, et al. J. Micromech. Microeng. 17 (2007) S1–S26

(111) (001)(110) →←

Pit-nucleation and Step-propagation

100 nm


Our concern

I th t i l t fl d l• Is the atomic scale step flow model applicable to macroscopic etching?pp p g

Etching solutions: KOH, TMAHEt hi d th t f i tEtching depth: tens of micrometers

• Does the number of dangling bonds determine the activeness of the surfaces ordetermine the activeness of the surfaces or steps?


Contents

Characterization of Anisotropic Etchingp gin macroscopic domains

“Dangling Bond Model” does not tell the truthDangling-Bond Model does not tell the truth,because no dynamics included.

“Step Flow Model” explains anisotropy in the vicinity of Si (111) y ( )

*Reversed anisotropy between KOH and TMAHReversed anisotropy between KOH and TMAH*Etched shape clearly reflects atomic-step

b h i i th i i it f Si (111)behavior in the vicinity of Si (111)


Etch Pit Growth on (111) Silicon.

Growth of individual pit traced during TMAH etching

K. Sato, et al.: Sensors and Materials 15-2 (2003) 93-99.

Growth of individual pit traced during TMAH etching.

10分後

Optical microscope images with time increments

・Wafer preparation… Oxidization 20分後 30分後p p

for 20 hours (3μm thick oxide) followed by oxide removal using HF.

20分後 30分後

・etching condition…25%TMAH, 80°30min


AFM image at the center of a pit etched with TMAH solution


Comparison in the shape of etch pits between KOH and TMAH

25%TMAH 80℃30分40%KOH 70℃30分[1 1 2][ 1 1 2]

K. Sato, et al.: Sensors and Materials 15-2 (2003)

[1 1-2][-1-1 2] [-1-1 2] [1 1-2]

Top View

100μmView

0.00

0.25

0.50

m)

m)

-0 4

-0.2

0.0

0.2

0.4

0.6

Cross-Section

0 75

-0.50

-0.25

tch p

it d

ept

h (

ｵm

Etching time: 10 min



Etc

h p

it d

epth

(ｵ

m

-1.4

-1.2

-1.0

-0.8

-0.6

0.4



0.250.200.150.100.050.00-1.25

-1.00

-0.75Et E

0.40.30.20.10.0-2.4

-2.2

-2.0

-1.8

-1.6 Etching time: 20 min



Distance (mm) Distance (mm)

Differently oriented etch pit growth governed by the difference in activated step

TMAHKOH TMAH


Difference in Etching Rate Contour Map between KOH and TMAH

(010)(010)90゜

(010)

(110)(111) 90゜

(010)

(110)(111)

60゜

(011)

60゜

30゜

(100)(001)

30゜

(100)(001)

0゜

0゜ 30゜ 60゜ 90゜

(100)

(101)

0゜0゜ 30゜ 60゜ 90゜

(100)

(101)

KOH40%70℃ TMAH25%80℃KOH40%70℃ TMAH25%80℃

min maxKOH: Etch rate sharply increases by misorientation toward [1 1-2]

TMAH: Etch rate sharply increases by misorientation toward [-1-1 2]


Contents

Characterization of Anisotropic Etchingin macroscopic domains

“Dangling-Bond Model” does not tell the truthDangling Bond Model does not tell the truth,because no dynamics included.

“St Fl M d l” l i i t i th“Step Flow Model” explains anisotropy in the vicinity of Si (111)

*Reversed anisotropy between KOH and TMAHReversed anisotropy between KOH and TMAH*Etched shape clearly reflects atomic-step

behavior in the vicinity of Si (111)behavior in the vicinity of Si (111)


Step Flow Model for (111) Silicon M. Elwenspoek, J. Electrochem. Soc. 140-7 (1993) 2075-80


Difference in surface structure depending on a direction of deviation from silicon (111)


Etched deep grooves on (110) Si using a KOH water solution

Groove depth: 90 microns


Asymmetric increase in etching rate by angular deviation from (111) orientation

(-11-1)

(110)

(-11-1)

(-112)


Answers to the questions

• Is the atomic scale step flow model applicable to macroscopic etching?to macroscopic etching?

YES, in the vicinity of (111)

Does the number of dangling bond determine• Does the number of dangling bond determine the activeness of the surfaces or steps?

NO(neither of the surface nor of steps)(neither of the surface, nor of steps)


Contents



“Step Flow Model” explains anisotropy in the vicinity of Si (111) c ty o S ( )




ReferencesEtch mechanisms and characterizationR.A. Wind and M.A. Hines, Surface Science 460 (2000) 21-38.M. Elwenspoek, J. Electrochem. Soc. 140-7 (1993) 2075-80.p , ( )M.A. Gosalvez, et al., J. Micromech. Microeng. 17 (2007) S1–S26.K. Sato, et al., Sensors and Actuators A-64 (1998) 87-93.K. Sato, et al., Sensors and Actuators A-73 (1999) 131-137.K. Sato, et al., Sensors and Materials 15-2 (2003) 93-99.

Effects of surfactantK. Sato, et al., Sensors and Materials 13-5 (2001) 285-291.Prem Pal, J. Micromech. Microeng. 17 (2007) 2299–2307.Prem Pal, et al, J. MEMS 18-6 (2009) 1345-1356.M.A. Gosalvez, et al., J. Micromech. and Microeng. 19-12 (2009) #125011.P P l t l J J A l Ph 49 (2010) 056702Prem Pal, et al., Jpn. J. Appl. Phys. 49 (2010) 056702.

Needle ArrayM. Shikida, et al., Proc. MEMS-03 (Kyoto, 2003), 562-565. M Shikida et al Sensors and Act ators A 116 (2004) 264 271M. Shikida, et al., Sensors and Actuators A 116 (2004) 264–271.M. Shikida, et al, J. Micromech. Microeng. 14 (2004) 1462–1467

Simulation systemK Sato et al Proc IFToMM Intl Micromechanism Symp (Tokyo 1993 6) 155 160K. Sato, et al, Proc. IFToMM Intl. Micromechanism Symp. (Tokyo, 1993.6) 155-160.K. Sato, et al., Electronics and Communications in Japan, Part2, 83-4 (2000). This

publication was translated from Trans. of the Inst. of Electronics, Information and Communication Engineers C-II J82-C-II no 3 (1999) 84-91


Communication Engineers C II, J82 C II, no. 3 (1999) 84 91.

basic 2 anisotropic wet-etching of silicon: characterization and

Documents